Can anybody think of a reason either my speedo or tach is slightly off when it's cold out and my car is still cold? Somehow it claims I'm doing 62mph at 2000rpms, when it can only do 59mph at 2000rpm. Once I get the interior warmed up (or it's warm outside), it's accurate. It's an electric analog speedo.

 

Analog gauges use a coil and magnet setup (a galvanometer) just like an analog multimeter. The drive electronics are temperature sensitive, how much depends on the sophistication of the circuit design and the parts used. Coil flux is also dependent on temp, so that could be an issue also. So, could be any of a few things.

It shouldn't be anything related to the sensors themselves, because they're usually digital and therefore aren't affected by temp unless they've failed.

 

So maybe it did this every winter and I just never noticed?

 

Yah, maybe. Could just be something in the electronics aging too, so it may not have done it forever.

I forgot to add that the electronics and galvanometer will self-heat when they're in use too, just from power dissipation.

 

Hmmm, this is a replacement speedo, but it had significantly lower miles than mine (don't worry internet police, I'm never selling the car but I rolled it forward to match my mileage anyways). I suppose I could swap over that coil pack, but it doesn't seem worth it.

 

I wouldn't worry about it. The car will spend most of it's time warm enough while you're driving that it shouldn't matter.

 

Can this kind of speedo be calibrated? It seems like even warmed up its about 1mph high, plus with these tires (when I say 1mph I mean compared to what the tach says assuming the stock tire size) even if it were correct it would be slightly off (so the 1mph is like 2.5mph actually), so maybe it's worth my paying someone to calibrate it.

Funny because these speedos traditionally read low, once you get into the triple digits anyways. When Motor Trend tested one back in the day it did 143mph at an indicated 138mph.

 

Many can be calibrated, I don't really know anything about which ones can or can't be. The calibration is done for a particular speed though, so you can't really make an analog gauge accurate over the entire measurement range. So if it's calibrated at 60MPH (just a random number) then it won't necessarily be accurate above or below that speed, but should be very close. That's probably why it was off 5MPH when Motortrend tested it. It could have been calibrated to be accurate at 140MPH, but then it wouldn't be accurate at 60MPH or whatever.
That's why digitally-driven, stepper motor gauges are so nice (and why almost all cars use them now AFAIK), the gauge is driven at little incremental positions (steps) instead of an analog current working against a return spring.

Anyway though, I wouldn't worry about it, if you know it's accurate (or accurate enough) then why worry about what it does before it warms up?

 

Isn't it slowly but surely affecting the accuracy of my odometer though? I think part of the problem is the old one was 1mph slow at 2000rpm (and I only recently discovered this when calculating what it really does at 2000rpm), so now reading a little bit fast seems so weird to me.